ABSTRACT
Background: Transcranial sonography (TCS) provides two-dimensional sonographic images
of the brain parenchyma through the intact skull. Objective: To determine the accuracy
and reliability of TCS and of a novel methodology for echogenicity quantification
in post hoc analysis of TCS images for the diagnosis of Parkinson’s disease (PD).
Method: 227 participants had standardized TCS examination. Gray scale mean (GSM) of
mesencephalic structures was determined from stored images. Results: 10.6% had no
bone windows. The groups included 77 PD patients and 126 controls. Substantia nigra
(SN) area diagnostic accuracy was 88.0% (85.7% sensibility, 89.7% specificity). For
SN > 0.23 cm2, area under the ROC curve = 0.925, Youden index = 0.75, likelihood ratios 8.32 positive
and 0.159 negative. The mesencephalic area did not discriminate the groups. Ventricular
diameters correlated positively with age and were higher in PD (p = 0.005). Hyperechogenic
foci in lentiform nucleus were scarce (2.5%). Reliability and reproducibility of measurements
were substantial. GSM showed superior values in Parkinsonians, with good reproducibility.
The two methodologies together identified all of the patients, yielding a 100% sensibility.
Conclusion: TCS is a reliable and accurate diagnostic tool for PD. The digital image
analysis proposed here further enhances its sensibility. TCS has good diagnostic value
besides its safety, convenience and low cost.
transcranial sonography - Parkinson’s disease - substantia nigra - diagnostic accuracy
- reproducibility
